goldensunfandomcom-20200223-history
Talk:Defend
I'm not going to simply re-edit in that deleted paragraph, that would feel like an edit war, lol, but I feel that a practical analysis does seem warranted in this article. Here is my analysis, re-phrased (since I did write that last paragraph quite badly). If others still feel this constitutes a "terrible understanding" of the game, I'm more than willing to hear any criticism, but in lieu of a valid counterargument, then I will put in back in eventually as there is nothing wrong with what I wrote (in principle at least), but I'll leave it here for right of reply first: :As a command, defend is almost always an undesirable option to use in battle, as a defending Adept cannot preform any other action and thus will not be able to aid the party on that turn. While the effect (reducing the rate of HP loss) is useful, defend is almost never the best choice, as increasing the rate of healing is almost always a much more effective tactic as healing items and Psynergy are plentiful and, with a few exceptions, will heal much more than 50% of an enemy's attack. However, for example, if the party cannot afford to loose any members and one or more Adepts do not have enough HP to survive a full hit, but the enemy acts first, then stalling until healing can be performed, by setting the endangered Adept(s) to defend, is potentially desirable. However, in this case, other defensive options (such as the Mars Djinni, Flash) are almost always better choices, as they offer greater protection and target more Adepts. Thus, the best time to use defend is when stalling is needed (it need not be waiting to restore HP, it could be Djinn recovery or status effects, such as Seal, etc), but all superior defensive options have already been consumed. Obviously, this is a niche situation that is both statistically and strategically unlikely to occur, and as such, it is far more likely that an Adept will simply be forced to use defend by default when an Adept's target in invalidated due to the monster being felled earlier in the turn. Slax01 00:00, November 28, 2009 (UTC) :I don't see how thats unreasonable to have, maybe give it a heading saying "analysis" or "strategy". Cheeseo is a little too condescending, but whatever, I think it's fine to add it back in personally. --Caasi 01:03, November 28, 2009 (UTC) ::I'd just like to note that something else I've always used defend for is when attempting to defeat single remaining enemies with the Dark Panther method of Djinn attacks to maximize EXP/coin/drop gains; Defend is useful for when you don't want your faster Jupiter adept to kill the enemy before your slower Mars Adept can unleash his offensive Mars Djinni on it, for example. So I think that should be integrated into there. Erik Jensen (Appreciate me here!) 02:06, November 28, 2009 (UTC) I deleted that paragraph because Defend has uses that it completely ignored. A good example is when you are fighting an enemy that is faster than you and the only way to prevent one of your characters from dying is to defend (the Briggs battle comes to mind). I'm not sure how the current version can give this exact example and also say that Defend is "almost always an undesirable option." The argument in the current revision that you can use defensive Djinni instead doesn't hold when you consider that a. you'll only one such Djinni in TLA (Shade) until Isaac joins late in the game, and won't even have that until Piers joins. b. only one character at a time can have it. c. it needs to be set. If you are lucky enough to have a defensive Djinni set at just the right time on the right character, then using it is probably a better option than defending. But to simply condemn the Defend command as useless like the old version of the article did does indeed show a poor understanding of the game. The current version helps people to draw their own conclusions despite the exaggerated ones present, so I won't change it except to get rid of the recurring "preform." Cheeseoman 04:28, December 2, 2009 (UTC) :Yeah sorry 'bout the recurring spelling errors, I'm converting Australian-> American as I type and I can never remember what words are different, so I often end up changing words that don't need to be changed, lol. But anyway, in counterargument; *while I'll agree it's useful against Briggs, that's simply because Briggs is a very tough boss if fought as early as possible, not because defend is a good move in of itself. That is to say; everything is useful when used against Briggs, because the player needs all the help they can get (assuming they don't just go get Piers first). Besides, I tried to put this in the first article, "While it is conceivable that the player may defend when... the Adept simply has to stall" by stall I meant "survive until the adept's turn", but I'm not so good at wordplay, so it probably reads poorly. Either way, the conclusion of nigh-uselessness is still valid, because of the following. *Consider this; defending twice is the same (damage intake) as not defending once. Also, the Adept must be on between 51% and 99% of the enemy's attack for defend to both render them unkilled and for a healing item to have any positive effect. Lower than 51% and they will die regardless. Higher than 100% and they will not die from the attack and be able to be healed without even using defend. This allows only a very narrow band of HP for the Adept to be on, which is why defend is statistically unlikely to be useful. Adding Shade to the mix only further decreases the probability, (but not much, because as you have correctly pointed out, there is a very good chance Shade will not be ready for use at the critical moment). Then take into account that letting the adept die and then casting revive will do to the party the same as defending (save for Djinn recover), though again, not by much, but it all adds up. *Second, take into account strategy. The best reason a party would get into a situation where some party members are injured more than the others (as at least one high-health Adept is needed to heal with) is if some Adepts have lower durability. Obviously, this is the case with Jupiter Adepts. However, they are also the ones with the greatest speed. If the enemy is acting before even the fastest member of you party (or the Jupiter Adepts have no healing items), then that implies the player is under-levelled or not assigning items well, which dictates that the player is either so good they don't need levels/items or so bad they can't level/have no items. Good will win regardless of what commands they choose and bad will loose regardless. Thus, the only players who stand to benefit are players who are "average". These players will generally not be under-levelled or itemless (by definition) and thus would almost always have a backup, such as slowing the enemy down or speeding their Adepts up and healing preventively to the point where they will almost never have the critical lack of speed and HP that would make defend useful. *Thirdly, take into account random chance. If your adept is in the critical threshold, there is almost always a chance that the enemy will simply not target that Adept, further reducing the probability of defend being actually useful, save for peace of mind. Sorry for the long-windedness (aren't you glad I didn't put that into the article, XD), but as you can see, there is a very low chance that the low speed-hp scenario will ever occur, and outside of that I see no use for defend (in a normal battle, it's very useful in Monster Battle), so I stand by my argument (though am more than willing to accept my wording sucks). Slax01 05:57, December 2, 2009 (UTC) Fair enough. Defend is certainly the least used command, but can be one of the most important when it's needed. So, I'm content as long as it's not portrayed as completely useless. I apologize if I came across as aggressive in my editing. Cheeseoman 17:28, December 2, 2009 (UTC) :No apologies needed at all, you were just trying to contribute, there's nothing wrong with that :) Slax01 22:35, December 2, 2009 (UTC) Automatic defense is not a true defense (Golden Sun and Golden Sun: The Lost Age) I put this in the talk because I am not sure about this and I find no source on the net to confirm it either. When I was analyzing the game's battle system in order to reproduce it by code, I realized that Defend does not apply exactly when the defend message is displayed. When a character Defends, he/she automatically reaches level 1 of barrier/protection (just an integer/enumeration value used by the game to represent the state during which all damage will be reduced by 50%) at the beginning of the round (turn for all characters). The defend message is then triggered when it is the defending character's turn (if it happens), but is just a notification. However the fact that the character does not skip his/her turn is important, since they may be some counters or other side effects on him/her (such as poisoning). This is what allows us to prevent some slow character from dying when he/she is running out of HP and an enemy faster than the healer (another character faster than the defender) attacks during the incoming turn. However, the automatic defence is different. First, it happens because a character was targetting a single or group of enemies/allies that disappeared due to previous actions during the same round (dying implies disappearing for enemies, but allies remain on the field so a cure will still be launched on a dead character, just ineffective). So the game does not predict this and does not trigger the barrier level 1 at the beginning of the round. However, if I remember well, when the character whose targets have died automatically defends and the defend message is displayed, the character's barrier level does not change, so again it is just a notification (but this time the barrier is at level 0 if no other changes occured) and the usual side-effects occur after that. In other words, the automatic defence is not really "smart" in the sense that it does not help the player by choosing an advantageous Defend move when the character has nothing to do, however it may be useful when you do not want your character to attack other enemies that the one targeted (some battle strategies require to temporarily keep some enemies alive; or you may want to kill some enemy with a Djinni of the appropriate element later for the bonus EXP), or do not want some spend some PP or Djinni for another enemy than the one targeted. When healing a dead character or resurrecting a character who is alive, however, the automatic "no action - defence" is not triggered so unfortunately you will consume/unleash some PP/Djinn for an uneffective action (having a Djinn in Standby can be useful for later though). In other words, the automatic Defend action is useful because it provides a high precision to the player's moves, not because it protects the characters better. What do you think? Can you confirm my assertion? (I have no emulator/GBA cartridge to check) Note: I used the term barrier/protection level. If level 1 stands for "normal defence", it is also triggered by level 1 barrier Djinn (Granite). Level 2 can be reached with Djinn Shade (damage -60%, or 40% damage taken), and level 3 with Djinn Flash (damage -90%, or 10% damage taken). The barrier level can only increase, not decrease, during the round (so combining several barriers simply keeps the effect the most powerful one), and is reset to 0 at the end of the round. Finally, if I remember well again, a dead character cannot benefit from a barrier and when he/she dies, he/she loses its previous barrier (like most temporary attributes, actually), and does not regain it when resurrected. That is why during critical battles where barriers are vital, characters that are resurrected with few HP at the beginning of middle of the round tend to die quickly just after that. So resurrecting them with a slow character is advised, since delaying a barrier is not possible, barrier-type Djinn having priority over other actions (strikes first, or actually a crazy augmentation of the priority value that should overcome any natural speed boost). Information source: tests on GBA emulator, reading and writing the RAM. Character stats are gathered at the same place, looking for their name or some stats in Hexadecimal for example should be enough. hs (no account) 14:26, August 10, 2014 (UTC)